QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lubin, F. Articles by Freedman, L. S. Search for Related Content PubMed PubMed Citation Articles by Lubin, F. Articles by Freedman, L. S. Pubmed/NCBI databases Substance via MeSH Medline Plus Health Information Diets Ovarian Cancer

---

Nutritional Epidemiology

Dietary Intake Changes and Their Association with Ovarian Cancer Risk¹

² Cancer Epidemiology Unit and ³ Biostatistics Unit, Gertner Institute, Sheba Medical Center, Tel Hashomer, Israel

^* To whom correspondence should be addressed. E-mail: floral{at}gertner.health.gov.il.

	ABSTRACT

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

 
There are reasons to suspect that dietary changes through adult life may modify risk for some cancers. We examined the association of recent and past dietary habits and changes in dietary intake over time with ovarian cancer risk. Long-term nutritional assessment was performed retrospectively in 631 incident cases of invasive epithelial ovarian cancer and in 1174 matched controls (matched by age ± 2 y, country of origin, and period of immigration) as part of a nationwide case-control study of ovarian cancer conducted between the years of 1994 and 1996 in Israel. Using a 2-step quantified Food Frequency Questionnaire, participants were first asked about their consumption of food items 1 y prior to the interview, and then whether their consumption had changed over time. The time of the change and consumption level before the change were recorded, allowing reconstruction of daily intakes for several time points. The reported mean percentage of animal fat intake decreased by 1.3% in cases but by 1.9% in controls (P for difference = 0.003). Conditional multivariate logistic regression was used to estimate odds ratios adjusted for total energy, parity, and oral contraceptive use. Substituting nonanimal fat in preference to animal fat over a relatively short term (between 2 and 7 y prior to interview) decreased the risk of ovarian cancer [OR = 0.65/100 kcal (418.4 kJ), 95% CI = 0.50 – 0.85]. Our results suggest that substitution of nonanimal for animal fat during adult life might reduce the risk of ovarian cancer, but this requires confirmation in prospective studies.

	Introduction

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

The maintenance of reproductive function is partly dependent upon nutrition and energy balance (2,3), and abnormalities in sexual maturation, menstrual cycles, and fertility are found to be related to overweight (4,5). Furthermore, ovarian cancer risk has been found to be positively associated with higher BMI in our study population (6) as well as in other ovarian cancer epidemiological studies (7–9), but not in all (10,11). Specific putative nutritional risk factors have been identified (12), in particular, animal fat intake. An international ecological study reported a positive correlation between the consumption of animal fat and mortality from ovarian cancer (13). Analytical studies evaluating diet as a risk factor for ovarian cancer have yielded contradictory results; whereas some failed to find an association (14,15), others found that an increased intake of animal fats were associated with an increased ovarian cancer risk (16–20). A meta-analysis of 6689 subjects from 8 observational studies that separately analyzed total fat, saturated and unsaturated fat, and animal fat, found that a high vs. low animal fat diet was associated with a raised relative risk (1.7) for ovarian cancer (21). Recently, a pooled analysis of 12 cohort studies of dietary fat and ovarian cancer reported a positive association for animal fat at the highest level of intake (with relative risk in fourth quartile 1.15; 95% CI 0.99–1.33) and ovarian cancer risk (22).

During adult life, people may dramatically change their dietary habits for reasons unrelated to the illness under investigation. These reasons include increased health consciousness, onset of chronic illness other than ovarian cancer, and importantly, in Israel, immigration. There are good reasons to believe that dietary changes during adult life may modify cancer risk, although this effect has only rarely been investigated. Studies show repeatedly that migrant populations who experience changes in lifestyle also experience changes in the incidence of cancer, compared with populations in their country of origin (23–25). Similar evidence from analytical studies are more difficult to find because most dietary assessments are limited to more recent intakes, such as those of the previous year. This is due, perhaps, to the difficulty of recording long-term diets and dietary changes over time.

We therefore developed a special methodology for case-control studies that records diet in the recent past (usually 1 y before interview), and all conscious changes that occurred before that time to 20 y back (26). In this study, we used this methodology to evaluate the role of past diet and dietary changes during adult life on ovarian cancer risk. Our hypothesis is that a diet higher in animal fat may promote the ovarian carcinogenesis process, so that reducing animal fat intake during a woman's adult life could inhibit this process and thus delay or prevent ovarian cancer.

	Materials and Methods

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

 
    Population. Nutritional data were collected for 631 consecutive incident cases of invasive epithelial ovarian cancer in Jewish Israeli women and 1174 matched controls. For most cases, 2 individually matched population controls were included, but for 81 of the cases, only 1 matched control was found. This sample was part of a larger comprehensive nationwide multidisciplinary case-control study conducted between 1994 and 1999 in Israel, which included 1707 cases and 3567 matched controls. Our sample comprises those cases entered during the years 1994–1996. In the remaining period, for reasons of cost, nutritional data were not collected. Continuous identification of all newly diagnosed ovarian cancer patients in Israel was conducted during the study period in all gynecological departments throughout the country. The population controls were individually matched to each case by age (±2 y), origin, place of residence, and period of immigration, and were randomly selected from the Israel National Population Registry (Table 1). Matching for origin and place of residence in Israel was useful in adjusting indirectly for socioeconomic status.

Our institution's Ethics Committee approved the study. Each subject gave written informed consent to participate. Multilingual, specially trained study interviewers conducted personal interviews. Cases were interviewed at the hospital, whereas controls were interviewed at home. Regular monthly monitoring of the interviewers' work was performed. In addition to information about nutrition, we also collected information concerning demographic variables, physical activity, reproductive history, endogenous and exogenous hormonal exposures, family history of cancer, past gynecological morbidity, drug use, radiation exposures, talc use, and other occupational chemical exposures.

    Dietary methods. The nutritional questionnaire was a 2-step quantified food frequency questionnaire (26) that we used previously in several other large case-control studies (27–31). This questionnaire provided the structure for recording a person's diet 1 y prior to the interview and changes in diet that were made as far back as 20 y before the interview. It included >180 food items, including traditional foods and alcoholic beverages, consumed by different ethnic groups in Israel. Respondents were asked to describe, for each food item, the frequency (daily, number of times/wk or mo, or never) and number of portions consumed per day as well as portion size. Natural units, standard units of packaging, and individual portion sizes, determined by life-sized color photographs, were used to quantify portion size. Reported food consumption was converted into mean daily amounts using a custom-built comprehensive computer program that contained information on food composition from international and local sources. Length of season was taken into consideration when calculating the mean daily consumption of seasonal fruits and vegetables.

The questionnaire was administered by an interviewer. First, respondents were asked a set of general introductory questions about any dietary changes that they had made during their adult life for reasons unrelated to ovarian cancer. Following this general overview, they were asked about their diet, item by item. First, they were asked about their recent consumption, that is, during the year prior to the interview (step 1). In addition, for each food item, the respondents were asked whether a change in habitual intake of that item had occurred during the last 20 y. If the answer was yes, the interviewer recorded the time (in years) since the change occurred, as well as the frequency and amount of consumption before the change (step 2).

This method of interview enabled reconstruction of the past dietary intake for each individual. For example, to construct an individual's diet 5 y prior to the interview, we used the present intake for all items except those reported to have changed within the 5 y previous to the interview. For the latter, we substituted the intake reported before the change. The same process was applied for 2 y, 7 y, and 20 y prior to the interview. To avoid including changes in dietary habits made in response to early symptoms of ovarian cancer, we did not analyze the "recent" (i.e., 1 y prior to the interview) dietary intake, but instead used the reconstructed intake 2 y prior to interview as our most recent one. This process allowed for an evaluation of past dietary intake as well as a description of the amounts and types of dietary changes that were made for reasons thought to be unrelated to ovarian cancer. The association of changes over time with ovarian cancer risk could then be assessed.

A detailed evaluation of our questionnaire has been published (26). Furthermore, the reliability of our interview method has been evaluated in previous epidemiological studies by reinterviewing subsamples of the population and has been found to be high. In our first case-control study using the same method, only 3 of 243 items included in the questionnaire had a reliability of <70% (27,28). Similar results were found in subsequent studies using the same methodology for assessing past and recent intakes (29,31).

    Statistical methods. Univariate analyses were first conducted comparing the intakes of cases and controls at 4 different time points (2 y, 5 y, 7 y, and 20 y previous to the interview). The differences were tested using univariate conditional logistic regression, with the matched sets as the stratifying variable. Because the results of these analyses (see Table 2) appeared strongly related to differences in energy intake, we also compared nutrient density intakes (energy intake of a specified nutrient expressed as a percentage of total energy) among cases and controls and their changes over the 4 time points mentioned above. (Results for 2 y and 7 y previous to interview are presented in Table 3).

The nonnutritional covariates were included in the model on the basis of previously established knowledge of the environmental risk factors for ovarian cancer. Thus, parity and oral contraceptive use were included as the most firmly established risk factors. Total energy intake was included in the model according to usual practice in nutritional epidemiology.

BMI was recorded in only 1566 of the 1805 women in the analysis. We therefore analyzed the data in 2 ways: for the 1566 women with BMI, we ran the models with BMI included as a covariate; for the total of 1805 women, we ran the models without BMI included as a covariate. The results for the nutritional variables were similar for both sets of analyses. We therefore chose to present the results for the analysis of the total group, without BMI as a covariate, because the extra numbers provide greater precision for the estimated OR.

Macronutrient intakes were entered as continuous variables into the models, and OR were expressed per 100 kcal of intake. This model is referred to in the literature as the "standard model" (33). The main reason for modeling continuous variables rather than quartiles of intake, for example, was because of our interest in associations with dietary change. Transformation of intakes into categories would complicate the interpretation of the statistical models that included dietary change.

In each analysis, total energy and 5 macronutrients were included, namely, carbohydrate, animal fat, nonanimal fat, animal protein, and nonanimal protein. However because the sum of intakes of these 5 macronutrients equals the total energy intake, they could not be entered simultaneously with total energy into a multivariate model; doing so would cause multicolinearity and would lead to unstable OR estimates. We therefore chose to omit animal fat.

The omission of animal fat from the model results in the interpretation that each macronutrient regression coefficient represents the effect of substituting that macronutrient for animal fat in the diet while maintaining the same total energy intake. This follows from the general rule in multiple regressions that a coefficient represents the effect on the outcome of changing the corresponding variable by 1 unit while keeping all other variables constant. We therefore considered the effect of changing the intake of a given macronutrient while maintaining the same total energy intake and the intake of all the other macronutrients included in the model. The only way to achieve this was to simultaneously reduce, by the same amount, the macronutrient omitted from the model, namely, animal fat.

Models were first run separately for the 4 different time points (2 y, 5 y, 7 y, and 20 y previous to the interview). The results of the joint model, including all 4 time points, indicated that the 5-y and the 20-y time points did not add to the risk model after including the 2-y and 7-y time points. Therefore, results were presented just for the model with 2-y and 7-y time points (see Table 4). Restriction to these 2 time points also simplifies the understanding of the results. Because we were primarily interested in the effect of change in the diet, we also reparameterized the model so as to analyze the change in intake between the 2-y and 7-y time periods, adjusted for intake at 7 y (results will be presented in Table 5).

Upon obtaining results indicating that a change in intake of animal fats was the main nutrient factor associated with ovarian cancer risk, we analyzed dietary intakes according to 24 food groups, concentrating on the groups that were the main sources of animal fat. Relative risks for change in food group intake were estimated using multivariate conditional logistic regression models similar to those used for the nutrient analyses.

	Results

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

 
Approximately 60% of the participants were born in Europe or America, and only 25% were <50 y of age (Table 1). Oral contraceptives were used by 20% of the control women. Only 6% of the control women were nulliparous. Among the cases, the main tumor histology was of the serous type.

The means and SD of intakes of various macronutrients were determined for cases and controls according to dietary intakes reconstructed at 2 y (i.e., recent), 5 y, 7 y, and 20 y previous to the interview (Table 2). For 97 participants <40 y of age at interview (34 cases and 63 controls), the "20-y previous" diet actually refers to their reported diet at age 20 y. Reported mean intakes of energy, protein, fat, and carbohydrates were all substantially higher among cases than controls in the 2 periods closest to the interview, but differed less in the 2 earlier periods. We also analyzed the change in intake over the period prior to interview in the 2 groups. Cases reported reducing their mean energy intake by 143 kcal (1954–1811) compared with 253 kcal in the controls during the period from 20 y to 2 y previous to the interview; during the period from 7 y to 2 y previous to the interview, the mean reported reductions were 55 kcal in cases and 136 kcal in controls.

Because the differences in nutrient intakes (Table 2) were at least partially confounded with the differences in reported energy intake of the cases and controls, we conducted all further analyses using some form of energy adjustment.

One simple ad hoc way of examining patterns adjusted for total energy intake is to consider each nutrient intake as a percentage of energy intake (nutrient densities). Animal fat densities decreased over time, whereas carbohydrate and nonanimal protein densities increased (Table 3). However, case subjects reported reducing animal fat density less often than controls. The mean percentage of energy intakes from animal fat at 7 y and 2 y previous to the interview was 19.9 and 18.6%, respectively among cases, and 20.0 and 18.1% among controls. Thus, over the 5-y period, the reported mean percentage of animal fat intake decreased by 1.3 in cases and by 1.9% in controls (univariate conditional logistic regression test for difference, P = 0.003). Also, the percentage of animal protein increased in controls, but was stable in cases (test for difference, P = 0.006, Table 3). Of particular note is the fact that the percentage of animal fat and animal protein did not differ between cases and controls at the individual time points, but their changes over time did (Table 3). A more comprehensive approach was performed using a multivariate analysis of energy-adjusted macronutrients at 2 y and 7 y previous to the interview (see Table 4). Animal fat was omitted from the model as described earlier. The interpretation of the OR for the remaining nutrients had the effect of substituting a given number (here, 100) of kcal from the nutrient for the same number of calories from animal fat. For 2 y previous to the interview, substitution of nonanimal fat for animal fat was associated marginally with a lower risk of ovarian cancer, whereas at 7 y previous to the interview there appeared to be no association (Table 4). Note that these analyses were adjusted for energy intake (through inclusion of energy intake in the model), parity, and oral contraceptive use.

The data suggest that change in dietary intake may be associated with ovarian cancer risk (see Table 2 and especially Table 3). We investigate this more directly and present a multivariate analysis of the change in intake adjusted for the baseline values (7 y previous to the interview) of several macronutrients over the 5-y period (from 7 y to 2 y previous to the interview; see Table 5). For intake changes between 7 y and 2 y previous to interview, the OR compared the ovarian cancer risk of 2 women who were similar in all respects except that one increased her intake of the nutrient (e.g., carbohydrate) by 100 kcal and reduced her intake of animal fat by 100 kcal over this period, whereas the other made no such changes. The results showed that increasing the levels of nonanimal fat for animal fat over the 5-y period was associated with a lower risk of ovarian cancer. There was also a suggestion that increasing carbohydrate and animal protein for animal fat may be similarly associated with a lower risk of ovarian cancer.

Having identified animal fats as the main "dietary-change" factor associated with ovarian cancer risk, we investigated which of the food groups might be involved. The major sources of animal fat in the study population were: dairy products, 37%; chicken or turkey, 21%; fish, 12%; beef, 7%; eggs, 7%; traditional food fats, 5%, and butter or margarine, 5%. A higher risk for ovarian cancer was found when 100 kcal of animal fat from eggs [OR 2.4, CI (1.1, 5.0)] or from chicken and turkey [OR 1.6, CI (1.0, 2.6)], were substituted for 100 kcal from nutrients that were not animal fat. Other sources of animal fat also estimated OR >1, but were not significant (P > 0.1) (dairy products, fish, beef, and butter/margarine).

To investigate how the substitution of animal protein for animal fat could be protective, we investigated the intake of the main source of animal fat, dairy products. Our data showed that the percentage of energy from animal fat in dairy products consumed by controls decreased from 47 to 43% over a period of 7 to 2 y from diagnosis, whereas, among cases, it decreased from 47 to 45%. In other words, the controls switched from higher fat to lower fat dairy products more often than the cases during this time period. The difference, though small, was significant (Z = 3.81, P < 0.001).

In sensitivity analyses excluding outliers (described in the Statistical Methods section), all foregoing associations did not change (Table 5), but the associations from substituting carbohydrate and animal protein for animal fat were not significant. However, substitution of nonanimal fat for animal fat remained associated with a lower risk of ovarian cancer (P < 0.05).

	Discussion

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

 
Although previous studies (13–22) have examined associations between dietary intake and ovarian cancer risk, to our knowledge, associations with changes in dietary intake have not previously been reported. In this study, we found that recent dietary habits were only marginally associated with ovarian cancer risk, and that the association of past dietary habits were even weaker. However, our evaluation of the association of changes in diet with disease indicated that increasing the substitution of nonanimal fat (and possibly carbohydrate and animal protein) for animal fat over a relatively short-term period (<7 y) was clearly associated with a decreased risk of ovarian cancer. We examined which food groups might be implicated in this substitution effect, and, although no one food group could be identified, we found stronger contributions from eggs and poultry. We examined how the substitution of animal proteins for animal fats could have a protective effect, and found that this effect might be explained by the fact that more women among the controls substituted low animal fat dairy products for high fat ones.

Dietary change in adult life is more often the rule than the exception. In a stratified random sample of the healthy adult Israeli population, we evaluated the prevalence of dietary change and showed that 68% of the population reported having changed the frequency of intake of at least 1 food item within the last 20 y, and that 38% reported changing >10 items. These reported changes tended to begin at the age of 40 y and increased with age. Among women, the mean change in energy intake from past to recent was 365 kcal (26), a little more than the mean reduction of 250 kcal reported by the controls in this ovarian cancer study. Trends toward a decrease in animal fat intake during adulthood have been reported in various European populations, with the effect especially marked among women (34–36).

Using the same dietary assessment method as in the European study, we found that disease associations differ according to whether one analyzes a "recent" diet or a diet 10 y previous to the interview in separate studies of the role of nutrition in stomach, colon, and breast cancer etiology (27–31). Past diet was generally more strongly associated with disease than recent diet. However, this is the first study in which we have explicitly evaluated dietary change as a risk factor.

Limitations inherent in the present study include a possible recall bias in past diets that may possibly be more marked when recalling changes over the long term. It is therefore important that our hypothesis be re-examined using data from large cohort studies where dietary change has been recorded. Another limitation in our analysis is the possibility of residual confounding with reported energy intake, which is itself strongly associated with disease in this study. The findings from the OPEN study by Subar et al. (32) suggest that such residual confounding could have an important effect; but it is unclear how far these results can be applied to a population who completed a rather different dietary questionnaire designed to assess not only recent but past dietary habits. Thus, residual confounding with energy cannot be demonstrated to be present nor can it be ruled out as an explanation of our results.

The strength of this study includes its large size, its use of population controls, the nationwide recruitment policy, the use of well-trained interviewers, the careful monitoring of their performance, and the use of questionnaires where the reliability of reported past and recent dietary habits was evaluated and found to be high (27,29,31). Moreover, our results are consistent with those of other epidemiological studies implicating animal fat as a dietary factor associated with ovarian cancer (13–22).

A possible advantage of studying dietary change, which deserves further exploration, is that changes in diet may be subject to less measurement error than absolute levels of dietary intake. Indeed this is one possible explanation for our finding of strong associations between dietary change and ovarian cancer and only weak associations with recent intakes.

We conclude that the consumption of animal fat in adult life is associated with an increased risk of ovarian cancer. In particular, our analysis suggests that ovarian cancer risk might be reduced by substituting energy from animal fat with fat from nonanimal sources or possibly carbohydrates and animal proteins. Our results, however, need to be confirmed in other epidemiological studies, particularly in prospective cohorts.

	ACKNOWLEDGMENTS

 
We thank S. Glickman, L. Shamir, E. Alfandary A. Zultan, H. Nitzan, and T. Rodkin.

	FOOTNOTES

 
¹ This research was supported by a grant from the National Cancer Institute (R01 CA61126-01-03).

⁴ Author is deceased.

⁵1 kcal = 4.18 kJ.

Manuscript received 30 November 2005. Initial review completed 14 February 2006. Revision accepted 12 June 2006.

	LITERATURE CITED

TOP ABSTRACT Introduction Materials and Methods Results Discussion LITERATURE CITED

 

1. Modan B, Hartge P, Hirsh-Yechezkel G, Chetrit A, Lubin F, Beller U, Ben-Baruch G, Fishman A, Menczer J, et al. Parity, oral contraceptives, and the risk of ovarian cancer among carriers and noncarriers of a BRCA1 or BRCA2 mutation. N Engl J Med. 2001;345:235–40.[Abstract/Free Full Text]

2. Basdevant A, Bringer JH, Lefebvre P. Weight, nutrition and hormonal events in women. Oxford University Press, Human Reproduction, 1997; 12 (Suppl 1): v–vii.

3. Simpson ER, Bulun SE, Nichols JE, Zhao Y. Estrogen biosynthesis in adipose tissue: regulation by paracrine and autocrine mechanisms. J Endocrinol. 1996;150: suppl:S51–7.[Medline]

4. Fishman J. Fatness, puberty and ovulation. N Engl J Med. 1980;303:42–3.[Medline]

5. Clark AM, Thornley B, Tomlinson L, Galletley C, Norman RJ. Weight loss in obese infertile women results in improvement in reproductive outcome for all forms of fertility treatment. Hum Reprod. 1998;13:1502–5.[Abstract/Free Full Text]

6. Lubin F, Chetrit A, Freedman LS, Alfandary E, Fishler Y, Nitzan H, Zultan A, Modan B. Body mass index at age 18 years and during adult life and ovarian cancer risk. Am J Epidemiol. 2003;157:113–20.[Abstract/Free Full Text]

7. Farrow DC, Weiss NS, Lyon JL, Daling JR. Association of obesity and ovarian cancer in a case-control study. Am J Epidemiol. 1989;129:1300–4.[Free Full Text]

8. Mori M, Nishida T, Sugiyama T, Komai K, Yakushiji M, Fukuda K, Tanaka T, Yokoyama M, Sugimori H. Anthropometric and other risk factors for ovarian cancer in a case-control study. Jpn J Cancer Res. 1998;89:246–53.

9. Tornberg SA, Carstensen JM. Relationship between Quetelet's index and cancer of the breast and female genital tract in 47,000 women followed for 25 years. Br J Cancer. 1994;69:358–61.[Medline]

10. Lukanova A, Toniolo P, Lundin E, Micheli A, Akhmedkhanov A, Muti P, Zeleniuch-Jacquotte A, Biessy C, Lenner P, et al. Body mass index in relation to ovarian cancer: a multi-center nested case-control study. Int J Cancer. 2002;99:603–8.[Medline]

11. Tornberg SA, Carstensen JM. Relationship between Quetelet's index and cancer of the breast and female genital tract in 47,000 women followed for 25 years. Br J Cancer. 1994;69:358–61.[Medline]

12. Huncharek M, Klassen Hm Kupelnick B. Dietary beta-carotene intake and the risk of epithelial ovarian cancer: a meta-analysis of 3,782 subjects from five observational studies. In Vivo. 2001;15:339–43.[Medline]

13. Rose DP, Boyan AP, Wynder EL. Interventional comparisons of mortality rates for cancer of the breast, ovary, prostate and colon and per capita consumption. Cancer. 1986;58:2363–71.[Medline]

14. Bertone ER, Rosner BA, Hunter DJ, Stampfer MJ, Speizer FE, Colditz GA, Willett WC, Hankinson SE. Dietary fat intake and ovarian cancer in a cohort of US women. Am J Epidemiol. 2002;156:22–31.[Abstract/Free Full Text]

15. Slattery ML, Schuman KL, West DW, French TK, Robinson LM. Nutrient intake and ovarian cancer. Am J Epidemiol. 1989;130:497–502.[Abstract/Free Full Text]

16. Bosetti C, Negri E, Franceschi S, Pelucchi C, Talamini R, Montella M, Conti E, La Vecchia C. Diet and ovarian cancer risk: a case-control study in Italy. Int J Cancer. 2001;93:911–5.[Medline]

17. Risch HA, Jain M, Marrett LD, Howe GR. Dietary fat intake and risk of epithelial ovarian cancer. J Natl Cancer Inst. 1994;86:1409–15.[Abstract/Free Full Text]

18. Zhang M, Yang ZY, Binns CW, Lee AH. Diet and ovarian cancer risk: a case-control study in China. Br J Cancer. 2002;86:712–7.[Medline]

19. Risch HA, Marrett LD, Jain M, Howe GR. Differences in risk factors for epithelial ovarian cancer by histologic type. Results of a case-control study. Am J Epidemiol. 1996;144:363–72.[Abstract/Free Full Text]

20. Kushi LH, Mink PJ, Folsom AR, Anderson KE, Zheng W, Lazovich D, Sellers TA. Prospective study of diet and ovarian cancer. Am J Epidemiol. 1999;149:21–31.[Abstract/Free Full Text]

21. Huncharek M, Kupelnick B. Dietary fat intake and risk of epithelial ovarian cancer: a meta-analysis of 6,689 subjects from 8 observational studies. Nutr Cancer. 2001;40:87–91.[Medline]

22. Genkinger JM, Hunter DJ, Spiegelman D, Anderson KE, Beeson WL, Buring JE, Colditz GA, Fraser GE, Freudenheim JL, et al. A pooled analysis of 12 cohort studies of dietary fat, cholesterol and egg intake and ovarian cancer. Cancer Causes Control. 2006;17:273–85.[Medline]

23. Ziegler RG, Hoover RN, Pike MC, Hildesheim A, Nomura AM, West DW, Wu-Williams AH, Kolonel LN, Horn-Ross PL, et al. Migration patterns and breast cancer risk in Asian-American women. J Natl Cancer Inst. 1993;85:1819–27.[Abstract/Free Full Text]

24. Parkin DM, Iscovich J. Risk of cancer in migrants and their descendents in Israel: II. Carcinomas and germ-cell tumors. Int J Cancer. 1997;70:654–60.[Medline]

25. Li XM, Ganmaa D, Sato A. The experience of Japan as a clue to the etiology of breast and ovarian cancers: relationship between death from both malignancies and dietary practices. Med Hypotheses. 2003;60:268–75.[Medline]

26. Lubin F, Chetrit A, Lusky A, Modan M. Methodology of a two-step quantified nutritional questionnaire and its effect on results. Nutr Cancer. 1998;30:78–82.[Medline]

27. Modan B, Lubin F, Barell V, Greenberg RA, Modan M, Graham S. The role of starches in the etiology of gastric cancer. Cancer. 1974;34:2087–92.[Medline]

28. Modan B, Barell V, Lubin F, Modan M, Greenberg A, Graham S. Low fiber intake as an etiological factor in cancer of the colon. J Natl Cancer Inst. 1975;55:15–8.[Medline]

29. Lubin F, Ron E, Wax Y, Black M, Furnaro M, Shitrit A. A case-control study of caffeine and methylxanthines in benign breast disease. JAMA. 1985;253:2388–92.[Abstract/Free Full Text]

30. Lubin F, Wax Y, Modan B. The role of fat, animal protein and dietary fiber in breast cancer etiology – a case-control study. J Natl Cancer Inst. 1986;77:605–12.[Medline]

31. Lubin F, Rozen P, Arieli B, Farbstein M, Knaani Y, Bat L, Farbstein H. Nutritional and lifestyle habits and water-fiber interaction in colorectal adenoma etiology. Cancer Epidemiol Biomarkers Prev. 1997;6:79–85.[Abstract]

32. Subar AF, Kipnis V, Troiano RP, Midthune D, Schoeller DA, Bingham S, Sharbaugh CO, Trabulsi J, Runswick S, et al. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: The OPEN Study. Am J Epidemiol. 2003;158:1–13.[Abstract/Free Full Text]

33. Kipnis V, Freedman LS, Brown CC, Hartman A, Schatzkin A, Wacholder S. Interpretation of energy adjustment models for nutritional epidemiology. Am J Epidemiol. 1993;137:1376–80.[Abstract/Free Full Text]

34. Waskiewicz A, Piotrowski W, Sygnowska E, Rywik S, Jasinski B. Did favourable trends in food consumption observed in the 1984–2001 period contribute to the decrease in cardiovascular mortality? Pol-MONICA Warsaw Project. Kardiol Pol. 2006;64:16–23.[Medline]

35. Krachler B, Eliasson MC, Johansson I, Hallmans G, Lindahl B. Trends in food intakes in Swedish adults 1986–1999: findings from the Northern Sweden MONICA (Monitoring of Trends and Determinants in Cardiovascular Disease) Study. Public Health Nutr. 2005;8:628–35.[Medline]

36. Marti-Henneberg C, Capdevila F, Arija V, Perez S, Cuco G, Vizmanos B, Fernandez-Ballart J. Energy density of the diet, food volume and energy intake by age and sex in a healthy population. Eur J Clin Nutr. 1999;53:421–8.[Medline]

This article has been cited by other articles:

				R. L. Prentice, C. A. Thomson, B. Caan, F. A. Hubbell, G. L. Anderson, S. A. A. Beresford, M. Pettinger, D. S. Lane, L. Lessin, S. Yasmeen, et al. Low-Fat Dietary Pattern and Cancer Incidence in the Women's Health Initiative Dietary Modification Randomized Controlled Trial J Natl Cancer Inst, October 17, 2007; 99(20): 1534 - 1543. [Abstract] [Full Text] [PDF]

				Minerva BMJ, September 2, 2006; 333(7566): 506 - 506. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lubin, F. Articles by Freedman, L. S. Search for Related Content PubMed PubMed Citation Articles by Lubin, F. Articles by Freedman, L. S. Pubmed/NCBI databases Substance via MeSH Medline Plus Health Information Diets Ovarian Cancer